How to measure your brand’s AI search visibility

Most brands can tell you where they rank on Google. Very few can tell you what ChatGPT says when a buyer asks it what to buy in their category.

That gap matters because AI answers do the comparing for the buyer and name only a handful of brands. If you are not measuring whether you are one of them, you cannot tell whether anything you do is working. This guide shows how to measure it, starting with nothing more than a spreadsheet.

What does AI search visibility mean?

AI search visibility is how often, and how accurately, AI assistants mention your brand when people ask questions you should be the answer to. It covers four things:

• Presence: is your brand named at all?
• Position: is it the first recommendation or the fifth?
• Accuracy: are the product facts, prices and positioning right?
• Citation: is your own site, or a page you influence, one of the sources the engine links to?

If you are new to the topic, start with what AEO is and how it differs from SEO.

How do you build the question set?

The question set is the foundation. Get it wrong and every number after it is noise.

1. Write 30 to 50 questions real buyers ask before purchasing in your category. Use your sales calls, customer service tickets, reviews and site search for the wording.
2. Cover four types: category questions (“best protein coffee for runners”), comparison questions (“Brand A vs Brand B”), fit questions (“best e-bike for a steep commute”) and objection questions (“is it worth the price”).
3. Include questions where you should NOT win, so you can see who the engines think owns them.
4. Freeze the list. Change it once a quarter at most. A measurement you keep changing cannot show a trend.

How do you run the measurement?

Ask every question in each engine, in a fresh chat with no history, and record the answer in a simple sheet.

Column	What to record
Question	The exact wording, unchanged every month
Engine	ChatGPT, Gemini, Perplexity, Google AI Overviews or AI Mode, Copilot
Named?	Yes or no
Position	1st, 2nd, 3rd mention, and so on
Description	How the engine described you, word for word
Competitors named	Every other brand in the answer
Sources cited	The domains the engine linked to

Run it on the same days each month. Answers vary from one run to the next, so for your most important questions, ask twice and record both.

What are the metrics that matter?

Four numbers summarize the sheet and are worth tracking monthly:

• Share of answer: the percentage of questions where your brand is named. This is the headline number.
• Share of first mention: the percentage where you are named first.
• Accuracy rate: the percentage of mentions with no factual errors.
• Citation share: the percentage of answers that cite your own domain.

Then compare each with your top three competitors. A share of answer of 20% means little alone. It means a lot if your nearest competitor sits at 60%.

The sources column is the most useful for action. It tells you which review sites, roundups, marketplaces and forums the engines lean on in your category, which is where to focus outreach.

How do you track AI referral traffic in Google Analytics?

Visibility should eventually turn into visits. In GA4, AI assistants show up as referral sources when a user clicks a cited link.

1. Go to Reports > Acquisition > Traffic acquisition and set the dimension to Session source.
2. Add a filter where session source matches the regex chatgpt|perplexity|gemini|copilot|claude.
3. Save it as a comparison or an exploration so you can check it weekly.

ChatGPT usually adds utm_source=chatgpt.com to the links it cites, which makes its traffic easier to spot. Treat these numbers as a floor: many people read the AI answer and then search for your brand directly, which shows up as branded search instead.

Should you use an AI visibility tool?

A growing set of paid trackers, such as Profound, Peec AI and Otterly.ai, automate the question runs and chart the results. They are worth it when your question set grows past what you can run by hand, or when you need daily data across many markets.

Start manually anyway. Running the questions yourself for the first month shows you how the engines think about your category in a way a dashboard will not. You will also write a better question set for the tool when you do buy one.

What should you do with the results?

Each gap in the sheet points to a fix:

• Not named at all: you lack the clear, answer-first pages and third-party mentions engines need. See how brands get recommended by ChatGPT, Gemini and Perplexity.
• Named, but described wrongly: your facts are inconsistent across your site, marketplaces and reviews. Make them match.
• Named, but never cited: your pages do not answer the question directly enough to quote. Rewrite them answer first.
• Competitor dominates one source: that site is your outreach target.

This baseline-and-fix cycle is the core of the Answer Layer work we run for brands. Our guide to the Answer Layer covers the full discipline.

Frequently asked questions

How often should you measure AI search visibility?

Monthly is enough for most brands. Weekly runs mostly capture random variation in the answers rather than real change.

Which AI engines should you track?

Start with ChatGPT, Google AI Overviews or AI Mode, Gemini and Perplexity. Add Copilot if your buyers are business users who work in Microsoft tools.

Why do AI answers change every time I ask?

AI assistants generate each answer fresh and may search the web live, so wording and brand lists vary between runs. That is why you track percentages across a fixed question set, not single answers.

Can I measure AI visibility for free?

Yes. A spreadsheet, a fixed question set and the free versions of the assistants are enough to start. Paid tools save time once you track many questions or markets.

What is a good share of answer?

There is no universal benchmark. Measure your own share against your top three competitors on the same questions, and aim to close the gap with the leader.